Session Item

SBRT treatment delivers a relatively high irradiation dose in a limited number of fractions and in a highly conformal manner, resulting in steep dose gradients between targets and OARs. Therefore, geometrical and anatomical variations can have a major impact on the planned dose distribution. Classical optimization considers uncertainties in patient setup, neglecting interfractional variations.     

Aim of the present study, side project of phase II randomized clinical trial RADIOSA (NCT03940235), is the development and testing of an anatomical robust optimization (aRO) method that can meet target coverage and dosimetric constraints despite variation in PTV position and bladder/rectal filling.

A total of 12 PCa oligometastatic pts with lymph node metastases enrolled within the RADIOSA trial, treated with 30Gy/3fx SBRT at the European Institute of Oncology and with isodose curve 50% intersecting bladder or/and rectum were considered.     
For each patient, a traditional plan was generated in the planning scenario (pCT) using 5 mm as PTV margin; a “MultipleCT-plan” was created using 3mm as CTV-to-PTV margin and by including 5 additional CTs (Figure 1), four simulating the shift of PTV by 5 mm in anterior, posterior, superior and inferior direction, and one (CT75) simulating an expansion of bladder and rectum by 0.24 and 0.27mm, respectively. These values represent the 75th percentile of collected variations of these OARs in a subgroup of RADIOSA-enrolled pts. Both plans simulated a 30Gy/3fx VMAT treatment optimized in Raystation v11.0 according to Timmerman constraints and doses  calculated with CCC algorithm. The two strategies were compared (i) on the nominal scenario (sCT) in terms of target coverage, PTV homogeneity index (HI), and constraints compliance; (ii) in terms of robustness, simulating a 0.52 mm shift of the patient in all the directions; (iii) on the CT75, in terms of variations in bladder/rectum D1.

Median value in the considered parameters on pCT and in the worst scenario (WS) across all pts can be seen in Table 1 for both strategies. Despite the smaller margin used, the proposed optimization method provided plans that can meet all constraints and clinical goals. In addition, at the robustness evaluation, aRO plans showed smaller variations between pCT and WS for all the considered parameters, pointing the increased robustness of the novel optimization method. Interestingly, HI variations resulted smaller for aRO plans, as well as rectum and bladder D1 variations at the dose recalculation on CT75.

These preliminary data suggest that inclusion of synthetic CTs as additional scenarios in the optimization process provides plans that can meet clinical goals despite the smaller PTV margins. In addition, aRO plans resulted more robust against changes in PTV position and anatomical variations, thus increasing the confidence in treatment delivery and potentially allowing for future dose escalation approaches.